#0: add host fallback when datatype is 4 or 8 bit which is not compat…

…ible with row major

ttnn/cpp/ttnn/operations/data_movement/reshape_view/reshape.cpp

@@ -55,9 +55,15 @@ ttnn::Tensor convert_tensor_to_rm_reshape_convert_back_to_orig_layout(const ttnn
     auto tensor_shape_with_padding = tensor_shape.padded_shape();
 
     //Constraint in device kernel
+
     uint32_t ROW_MAJOR_WIDTH = 32/tensor.element_size();
     ttnn::Tensor reshaped_rm_tensor;
-    if((tensor_shape[-1] % ROW_MAJOR_WIDTH == 0 && shape[-1] % ROW_MAJOR_WIDTH == 0)) {
+    if(tensor.element_size()<2 && layout == ttnn::TILE_LAYOUT)
+    {
+        //Can't call to_layout on 4b and 8b datatypes
+        reshaped_rm_tensor = host_reshape(tensor, shape);
+    }
+    else if((tensor_shape[-1] % ROW_MAJOR_WIDTH == 0 && shape[-1] % ROW_MAJOR_WIDTH == 0)) {
         auto rm_tensor = ttnn::to_layout(tensor, ttnn::ROW_MAJOR_LAYOUT, std::nullopt, std::nullopt, (Device *)nullptr);
         if (rm_tensor.is_contiguous()) {
             // Page size depends on the width, so only modify the shape if the width is the same
@@ -105,11 +111,13 @@ ttnn::Shape tiling_reshape_corrector(const ttnn::Shape& shape) {
     auto padded = shape.with_tile_padding();
     auto rank = shape.rank();
     const int8_t correction_1 =(ttnn::types::TILE_SIZE - (int)padded[-1] % ttnn::types::TILE_SIZE) % ttnn::types::TILE_SIZE;
+    if(rank == 1)
+    {
+        return ttnn::Shape({shape[0]},{padded[0]+correction_1});
+    }
     const int8_t correction_2 =(ttnn::types::TILE_SIZE - (int)padded[-2] % ttnn::types::TILE_SIZE) % ttnn::types::TILE_SIZE;
     switch(rank)
     {
-        case 1:
-            return ttnn::Shape({shape[0]},{padded[0]+correction_1});
         case 2:
             return ttnn::Shape({shape[0],shape[1]},{padded[0]+correction_2,padded[1]+correction_1});
             break;